摘要:
One aspect of the invention is a method for altering defects in a digital image. At least a first pixel of a first channel of a digital image is filtered using digital circuitry to produce a filtered pixel by averaging the intensity of the first pixel and a plurality of additional pixels in the neighborhood of the first pixel. At least one of the pixels is weighted in response to the intensity value of at least one pixel in a defect channel associated with the digital image. A corrected digital image is produced using the digital circuitry in response to the filtered pixel and the first channel of the digital image.
摘要:
During electronic film development, an area of conventional photographic film is scanned several times using a single scanning station, and at each subsequent time this scanned area is advanced incrementally along the film with multiple levels of overlap with previous scans. The new image scanned at each new time is aligned to an accumulating image that has been extrapolated to the image at the new time, and then the new image is added to the accumulating image in parametric summations that allow an image to be interpolated to any time free of seams where the scans overlap. The invention further teaches a method of steering the alignment by warping the leading edge of the alignment, and a registration method of aligning multiple images that takes advantage of known fixed alignments between images.
摘要:
Surface defect correction technology for photographic images requires an infrared scan along with a conventional color scan. In the present invention, the additional infrared scan needed for surface defect correction is obtained by adding a line of sensors specific to infrared light to a conventional multilinear color sensor array. The invention teaches a practical mode of distinguishing infrared light using a dichroic prism placed over the sensor. This mode has the additional advantage of placing the infrared-specific sensor line in a displaced focus plane to match conventional lenses. Adding a sensor line to a conventional trilinear sensor array requires a quadrilinear array topology. In addition to the direct quadrilinear topology, the invention teaches a method of obtaining full color image information with only two linear sensor lines by interstitially mixing red and blue sensors on a single sensor line, which, in conjunction with the additional infrared line, results in a conventional trilinear sensor topology with a different filter arrangement.
摘要:
One aspect of the invention is a method for normalizing a plurality of signals wherein the plurality of signals have a shared component and wherein at least one of the signals has been distorted in a nonlinear way. A distortion function is determined for at least one of the signals which is proportional to the distortion of that signal relative to at least one of the remaining signals. An inverse relative distortion function is generated for the distorted signal responsive to the distortion function that was determined for that signal. The signal is normalized by applying the inverse relative distortion function that was generated for the distorted signal.
摘要:
Surface defects in a reflection scan of a print made with visible light are corrected by using a scan of the print made with infrared light. In accordance with the present invention, surface defects in a reflection scan of an image consisting of pixels made with visible light are corrected by using a scan of the image consisting of pixels made with infrared light. This correction of surface defects is performed by first establishing for each pixel an upper and lower bound for defect intensity based on the infrared record. The corresponding visible pixel is then corrected by subtracting the combination of upper and lower bound resulting in a corrected pixel.
摘要:
An image for digital image processing is often parsed into tiles, then the processing is applied to each tile separately. This segmentation reduces active memory and processing time compared to that needed to process the entire image as a whole. The image is parsed into a larger number of overlapping tiles. After image processing, the blocks are blended by multiplying each with a blending window chosen to be mutually constant, and the blocks are added together to form the product image free of transition artifacts. In one implementation the centers of the overlapping blocks are on a diagonal grid. In another implementation, a function of the complete window, typically the square root of each pixel of the window, is applied to the blocks prior to processing and the remainder applied after processing. This softens tile edges going into the image processing application to permit better results from edge sensitive functions, such as a Fourier transform.
摘要:
Recovering the dye image on film in electronic film development following a latent holding stage obviates the problem common in prior art electronic film development of film image destruction. Recovery of the image is accomplished using a developing agent containing couplers to form a dye image. Theses dyes do not affect the infrared scans of the image. Upon complete development of the dye image, further dye formation is halted by the application of a coupler blocking agent, while silver development and electronic scanning may continue or halt. After halting dye formation, the film is stable for an arbitrary time in a latent stage and may be dried and stored. Following this latent stage, silver is removed from the film with a bleach-fix leaving a conventionally usable film image.
摘要:
During electronic film development, an area of conventional photographic film is scanned several times using a single scanning station, and at each subsequent time this scanned area is advanced incrementally along the film with multiple levels of overlap with previous scans. The new image scanned at each new time is aligned to an accumulating image that has been extrapolated to the image at the new time, and then the new image is added to the accumulating image in parametric summations that allow an image to be interpolated to any time free of seams where the scans overlap. The invention further teaches a method of steering the alignment by warping the leading edge of the alignment, and a registration method of aligning multiple images that takes advantage of known fixed alignments between images.
摘要:
During electronic film development, convential film is scanned electronically to produce a series of views of the develing image. An early scan reveals the fast developing highlight detail, while a late scan reveals slow developing shadow detail. After development, the series of views is combined into a single image in a process called stitching. In the present invention, regression data is accumulated during development to describe a curve of density versus time of development for each pixel. After development, this regression data is used to recreate a regression curve of dentisity versus development time for each pixel. The time at which this curve crosses a density known to give optimum grain characteristics, called the optimum density curve, is used to create the brightness for that pixel in the finished stitched image. The invention further teaches weighting regression data as a function of time and density generally following proximity to optitum density curve.
摘要:
A single pass scanner having a trilinear array (100), a source of white light (110), filters of the three primary colors and a separate source of infrared light is used in various methods of removing medium-based defects from a scanned film image. The method generates an infrared channel in addition to the common visable channels by covering the parallel rows of sensors in the trilinear array respectively with a red, green and blue filter to create the three color channels. Normally, each of the three color filters also passes infrared light, which is removed by filters external to the sensors. In a specific embodiment, interstitial in time between two visible light scans, the sensor is exposed to infrared light for a single scan. As the trilinear array sweeps across an image in time and spatial synchronization with the exposing lights, at least two visable channels and an infrared channel are generated.